期刊,Robotics & Machine Learning Daily News 2024年卷Mar.29期_国家学术搜索

期刊信息/Journal information

Robotics & Machine Learning Daily News

NewsRx

主办单位：NewsRx

Robotics & Machine Learning Daily News/Journal Robotics & Machine Learning Daily News

正式出版

收录年代

Findings on Robotics Reported by Investigators at Karlsruhe Institute of Technol ogy (KIT) (Uncertainty-aware Hand-eye Calibration)

1-1页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Research findings on Robotics are disc ussed in a new report. According to news reporting originating in Karlsruhe,Ger many,by NewsRx journalists,research stated,"We provide a generic framework fo r the hand-eye calibration of vision-guided industrial robots. In contrast to tr aditional methods,we explicitly model the uncertainty of the robot in a stochas tically founded way." The news reporters obtained a quote from the research from the Karlsruhe Institu te of Technology (KIT),"Albeit the repeatability of modern industrial robots is high,their absolute accuracy typically is much lower. This uncertainty-especia lly if not considered-deteriorates the result of the hand-eye calibration. Our p roposed framework does not only result in a high accuracy of the computed hand-e ye pose but also provides reliable information about the uncertainty of the robo t. It further provides corrected robot poses for a convenient and inexpensive ro bot calibration. Our framework is computationally efficient and generic in sever al regards. It supports the use of a calibration target as well as self-calibrat ion without the need for known 3-D points. It optionally enables the simultaneou s calibration of the interior camera parameters. The framework is also generic w ith regard to the robot type and,hence,supports antropomorphic as well as sele ctive compliance assembly robot arm (SCARA) robots,for example. Simulated and r eal experiments show the validity of the proposed methods."

原文链接:

NETL
NSTL

Data from University of California San Diego (UCSD) Provide New Insights into Ro botics and Automation (Shenron - Scalable,High Fidelity and Efficient Radar Sim ulation)

2-2页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators discuss new findings in Robotics - Robotics and Automation. According to news reporting from La Jolla,C alifornia,by NewsRx journalists,research stated,"Radar Simulations have becom e an essential tool in radar algorithm development and testing due to the lack o f available high-resolution radar datasets and enormous difficulty in acquiring real-world data. However,simulating radar data is challenging as existing radar simulation tools are not easily accessible,require detailed mesh inputs and ta ke hours to simulate." The news correspondents obtained a quote from the research from the University o f California San Diego (UCSD),"To address these issues,we present SHENRON,an open-source framework that efficiently simulates high-fidelity MIMO radar data u sing only lidar point cloud and camera images. We show that with SHENRON,one ca n generate simulated data that can be used to evaluate algorithms as effectively as on real data."

原文链接:

NETL
NSTL

Studies from China University of Petroleum in the Area of Machine Learning Repor ted (Explainable Machine-learning Predictions for Catalysts In Co2-assisted Prop ane Oxidative Dehydrogenation)

2-3页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators publish new report on Ma chine Learning. According to news originating from Beijing,People's Republic of China,by NewsRx correspondents,research stated,"Propylene is an important ra w material in the chemical industry that needs new routes for its production to meet the demand. The CO2-assisted oxidative dehydrogenation of propane (CO2-ODHP ) represents an ideal way to produce propylene and uses the greenhouse gas CO2." Financial supporters for this research include State Key Laboratory of Heavy Oil Processing,SINOPEC Petroleum Exploration and Production Research Institute. Our news journalists obtained a quote from the research from the China Universit y of Petroleum,"The design of catalysts with high efficiency is crucial in CO2- ODHP research. Data-driven machine learning is currently of great interest and g aining popularity in the heterogeneous catalysis field for guiding catalyst deve lopment. In this study,the reaction results of CO2-ODHP reported in the literat ure are combined and analyzed with varied machine learning algorithms such as ar tificial neural network (ANN),k-nearest neighbors (KNN),support vector regress ion (SVR) and random forest regression (RF)and were used to predict the propylen e space-time yield. Specifically,the RF method serves as a superior performing algorithm for propane conversion and propylene selectivity prediction,and SHapl ey Additive exPlanations (SHAP) based on the Shapley value performs fine model i nterpretation. Reaction conditions and chemical components show different impact s on catalytic performance."

原文链接:

NETL
NSTL

New Robotics Study Findings Recently Were Reported by Researchers at Hebei Unive rsity of Technology (Design and Characterization of Large-range 3-d Force Tactil e Sensor Based On Fe-ga Alloy)

3-4页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators publish new report on Ro botics. According to news reporting originating from Tianjin,People's Republic of China,by NewsRx correspondents,research stated,"Tactile sensors are among the essential components for human-robot interaction in robotics. The force appl ied to the robot hand is coupled with both normal force (in the z-direction) and shear force (in the x- and y-directions)." Financial support for this research came from National Natural Science Foundatio n of China (NSFC). Our news editors obtained a quote from the research from the Hebei University of Technology,"Additionally,the robot hand must withstand significant force whil e grasping the object. As a result,it is crucial to develop a sensor that can m easure a wide range of 3-D forces. This article proposes a tactile sensor for 3- D force based on Fe-Ga alloy and hyperelastic material polydimethylsiloxane (PDM S). The output voltage model of the Fe-Ga alloy sensing unit has been establishe d. The tunneling magnetoresistance sensor outputs a voltage signal that reflects changes in the magnetic field. We established a 3-D force simulation test platf orm to examine the static and dynamic characteristics of the sensor. The experim ental results indicate that the sensing unit's sensitivity is 310.15 mV/N betwee n 0-and 5-N range,and the experimental output voltage matches the calculated va lue. The sensor measured the normal force between 0-and 25-N range with a sensit ivity of 56.53 mV/N and the shear force between 0-and 12.5-N range with a sensit ivity of 50.82 mV/N. The sensor's response time and recovery time,which are les s than the human skin response time,are 40 and 42 ms,respectively."

原文链接:

NETL
NSTL

Research from Chinese Academy of Sciences Has Provided New Data on Machine Learn ing (Machine Learning for Chemistry: Basics and Applications)

4-5页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New study results on artificial intell igence have been published. According to news reporting out of Shanghai,People' s Republic of China,by NewsRx editors,research stated,"The past decade has se en a sharp increase in machine learning (ML) applications in scientific research ." Our news journalists obtained a quote from the research from Chinese Academy of Sciences: "This review introduces the basic constituents of ML,including databa ses,features,and algorithms,and highlights a few important achievements in ch emistry that have been aided by ML techniques. The described databases include s ome of the most popular chemical databases for molecules and materials obtained from either experiments or computational calculations. Important two-dimensional (2D) and three-dimensional (3D) features representing the chemical environment of molecules and solids are briefly introduced. Decision tree and deep learning neural network algorithms are overviewed to emphasize their frameworks and typic al application scenarios. Three important fields of ML in chemistry are discussed: retrosynthesis,in which ML predicts the likely routes of organic synthesis; atomic simulations,which utilize the ML potential to accelerate potential energ y surface sampling; and heterogeneous catalysis,in which ML assists in various aspects of catalytic design,ranging from synthetic condition optimization to re action mechanism exploration."

原文链接:

NETL
NSTL

Reports on Robotics Findings from Shanghai Jiao Tong University Provide New Insi ghts (Attentionvote: a Coarse-to-fine Voting Network of Anchor-free 6d Pose Esti mation On Point Cloud for Robotic Bin-picking Application)

5-6页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Current study results on Robotics have been published. According to news reporting originating from Shanghai,People's Republic of China,by NewsRx correspondents,research stated,"Current state-of -the-art pose estimation methods are almost launched on segmented RGB-D images. However,these methods may not apply to more general industrial parts due to a l ack of texture information and highocclusion of stacked objects." Financial support for this research came from National Natural Science Foundatio n of China (NSFC). Our news editors obtained a quote from the research from Shanghai Jiao Tong Univ ersity,"This article establishes an end-to-end pipeline to synchronously regres s all potential object poses from an unsegmented point cloud. The point pair fea tures (PPFs) are first extracted and then fed into a PointNet-like backbone for obtaining the point-wise features. Based on the center voting,a coarse-to-fine voting architecture is proposed to extract instance features instead of implemen ting instance segmentation. A lightweight threedimensional (3D) heatmap is leve raged to cluster votes and generate center seeds. Further,an attention voting m odule is constructed to fuse point-wise features into instance-wise features ada ptively. Ultimately,the suggested network regresses object poses with a quatern ion loss to handle the symmetric puzzle. The network holds the advantage of prod ucing the final pose prediction without any post-processing steps like nonmaximu m suppression (NMS) or any pose refinement modules like iterative closest point (ICP). The proposed network is evaluated on the public Fraunhofer IPA dataset,w hich demonstrates the robustness of the pose estimation network with much better performance."

原文链接:

NETL
NSTL

Vita-Salute San Raffaele University Reports Findings in Artificial Intelligence (Assessing Diabetic Retinopathy Staging With AI: A Comparative Analysis Between Pseudocolor and LED Imaging)

6-7页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New research on Artificial Intelligenc e is the subject of a report. According to news reporting originating in Milan,Italy,by NewsRx journalists,research stated,"To compare the diagnostic perfor mance of artificial intelligence (AI)-based diabetic retinopathy (DR) staging sy stem across pseudocolor,simulated white light (SWL),and light-emitting diode ( LED) camera imaging modalities. A cross-sectional investigation involved patient s with diabetes undergoing imaging with an iCare DRSplus confocal LED camera and an Optos confocal,ultra-widefield pseudocolor camera,with and without SWL." The news reporters obtained a quote from the research from Vita-Salute San Raffa ele University,"Macula-centered and optic nerve-centered 45 x 45-degree photogr aphs were processed using EyeArt v2.1. Human graders established the ground trut h (GT) for DR severity on dilated fundus exams. Sensitivity and weighted Cohen's weighted kappa (wk) were calculated. An ordinal generalized linear mixed model identified factors influencing accurate DR staging. The study included 362 eyes from 189 patients. The LED camera excelled in identifying sight-threatening DR s tages (sensitivity = 0.83,specificity = 0.95 for proliferative DR) and had the highest agreement with the GT (wk = 0.71). The addition of SWL to pseudocolor im aging resulted in decreased performance (sensitivity = 0.33,specificity = 0.98 for proliferative DR; wk = 0.55). Peripheral lesions reduced the likelihood of b eing staged in the same or higher DR category by 80% (P <0.001). Pseudocolor and LED cameras,although proficient,demonstrated noninte rchangeable performance,with the LED camera exhibiting superior accuracy in ide ntifying advanced DR stages. These findings underscore the importance of impleme nting AI systems trained for ultra-widefield imaging,considering the impact of peripheral lesions on correct DR staging."

原文链接:

NETL
NSTL

Researcher at Istanbul Technical University Has Published New Data on Robotics (Indoor 2D Autonomous Exploration with an Omnidirectional Robot: A Strategy Based on Rapidly-Exploring Random Trees)

7-8页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators publish new report on ro botics. According to news reporting originating from Istanbul,Turkey,by NewsRx correspondents,research stated,"An essential characteristic of a fully autono mous robot is the capability to examine an unfamiliar environment and construct a representation of it." The news reporters obtained a quote from the research from Istanbul Technical Un iversity: "The challenge of autonomous exploration involves overcoming various s ub-problems,including Simultaneous Localization and Mapping (SLAM),motion plan ning,target identification,and informed decision-making for target selection. This paper presents a frontier-based methodology to identify potential navigatio n targets for the autonomous exploration of unknown environments by an omnidirec tional robot. Permanent and temporary Rapidly-exploring Random Tree (RRT)-based structures are used to search the map and detect frontier points. A novel temporary RRT-based structure,Frontier Temporary Tree,is introduced in this study. It is noteworthy that RRT is solely used to search the explorable space for front ier points and does not contribute to motion planning. A cost-benefit analysis,taking into account path cost,heading cost,and information gain,is used to ev aluate the frontier points and determine the best target among them."

原文链接:

NETL
NSTL

New Findings on Robotics Described by Investigators at Oregon State University (The Effect of Uneven and Obstructed Site Layouts In Best-of-n)

8-9页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators publish new report on Ro botics. According to news originating from Corvallis,Oregon,by NewsRx correspo ndents,research stated,"Biologically inspired collective decisionmaking algor ithms show promise for implementing spatially distributed searching tasks with r obotic systems. One example is the best-of-N problem in which a collective must search an environment for an unknown number of sites and select the best option. " Financial support for this research came from Office of Naval Research. Our news journalists obtained a quote from the research from Oregon State Univer sity,"Real-world robotic deployments must achieve acceptable success rates and execution times across a wide variety of environmental conditions,a property kn own as resilience. Existing experiments for the best-of-N problem have not expli citly examined how the site layout affects a collective's performance and resili ence. Two novel resilience metrics are used to compare algorithmic performance a nd resilience between evenly distributed,obstructed,or unobstructed uneven sit e configurations. Obstructing the highest valued site negatively affected select ion accuracy for both algorithms,while uneven site distribution had no effect o n either algorithm's resilience."

原文链接:

NETL
NSTL

Cairo University Reports Findings in Artificial Intelligence (Artificial intelli gence system for automatic maxillary sinus segmentation on cone beam computed to mography images)

9-10页

查看更多>>摘要：By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New research on Artificial Intelligenc e is the subject of a report. According to news reporting originating from Cairo ,Egypt,by NewsRx correspondents,research stated,"The study aims to develop a n artificial intelligence (AI) model based on nnU-Net v2 for automatic maxillary sinus (MS) segmentation in Cone Beam Computed Tomography (CBCT) volumes and to evaluate the performance of this model. In 101 CBCT scans,MS were annotated usi ng the CranioCatch labelling software (Eskisehir,Turkey) The dataset was divide d into three parts: 80 CBCT scans for training the model,11 CBCT scans for mode l validation,and 10 CBCT scans for testing the model." Our news editors obtained a quote from the research from Cairo University,"The model training was conducted using the nnU-Net v2 deep learning model with a lea rning rate of 0.00001 for 1000 epochs. The performance of the model to automatic ally segment the MS on CBCT scans was assessed by several parameters,including F1-score,accuracy,sensitivity,precision,Area Under Curve (AUC),Dice Coeffic ient (DC),95% Hausdorff Distance (95% HD),and Inte rsection over Union (IoU) values. F1-score,accuracy,sensitivity,precision val ues were found to be 0.96,0.99,0.96,0.96 respectively for the successful segm entation of maxillary sinus in CBCT images. AUC,DC,95% HD,IoU v alues were 0.97,0.96,1.19,0.93,respectively."

原文链接:

NETL
NSTL